Several optical absorption measurement techniques have been proposed for gaseous species monitoring and quantification. One technique irradiates a sample with a broadband frequency source and disperses the transmitted light with a device such as a diffraction grating. The dispersed light is then monitored for the presence of a characteristic absorption line of the species of interest. However, in gas mixtures the absorption lines of the different species are often closely spaced. If they are, it is extremely difficult to disperse the light finely enough to have sufficient resolution to detect the single line of interest. Thus, the dispersal technique does not allow the monitoring of many gas mixtures, which prevents the dispersal monitors from having universal application. Even if the absorption line can be distinguished, the device used to disperse the light is typically complex and costly. Thus, optical gaseous species monitors employing radiation dispersal techniques have limited utility, and are typically very costly.
Some of the problems created by the use of a broadband source have been alleviated by the substitution of a narrow band laser source. Single or multiple mode semiconductor diode lasers having bandwidths which may be approximately the same width as an absorption line have been employed as the radiation source in optical absorption species monitors. One such technique is disclosed in U.S. Pat. No. 4,730,112. The wavelength of the source is typically slowly scanned by varying the laser drive current through a range of wavelengths including the absorption line of interest. The absorption line is then found by the detection of an intensity dip in the transmitted radiation. The species concentration may be determined by measuring the intensity decrease, which may be accomplished by comparing the amount of incident radiation to the amount of transmitted radiation; the greater the absorption, the greater the species concentration. However, this technique does not disclose any means for compensating for variations in absorption depth due to changes in line width associated with pressure and temperature variations. Thus, the reference apparatus has relatively limited utility.
The patented technique has three inherent difficulties which are typical of such scanning techniques. First, it is difficult to identify the absorption of interest and determine its exact magnitude when other absorption features are present in the spectral scan. Therefore such techniques have limited application in cases where there are other possible interfering absorptions within the spectral scan. Second, because the laser must be scanned over a large wavelength range, the sample species absorption is only observed a small fraction of the time. Therefore, detection sensitivity in the narrow band containing the absorption of interest and capabilities for real time monitoring of the sample species are limited. Only changes which occur on a time scale much longer than that of a single scan can be detected. Third, most available laser sources have instabilities in the output wavelength which result in spurius absorption measurements and distort the absorption line shape. These limit the accuracy of a scanning measurement. For gallium aluminum arsenide lasers such as those used in the patented device, these instabilities limit detection sensitivity in a scanning mode.
Because of the inherent difficulties, the scanning technique is suitable only for a limited set of sample species and conditions, such as oxygen monitoring at atmospheric pressure. Application to a larger range of monitoring conditions requires a technique which can accurately determine sample concentrations for weak absorbers with variable absorption line widths; the scanning technique measures only the absorption strength at the line center, and does not include a method to determine concentration in cases where pressure or temperature variations can alter the absorption line width and peak absorption strength.